Sunday, December 31, 2023

Changeability

It's pretty well known around here that I come down pretty heavily on the side of genetics in any "Why do people act like they do?" discussions. (Note: I do make allowances for both incentives and serious environmental issues like head injury, substance abuse, and trauma.) The Unsupervised Learning podcast, though run by a geneticist, has put up yet another challenge to this, an interview with Brent Roberts, psychology prof at U Illinois (AI transcript here*). The other recent challenge was by Rob Henderson, which I wrote about in August. Shortest version: Environment matters as long as they are living in your environment, but genetics asserts itself when they leave the nest, job, military, country, etc.

Roberts makes a very good case for his opinion that changeability is the key question, more than environmental "hopium" as he calls it, of foundations and public officials investing in two week boot camps, a few sessions of affirmation speech in the schools to optimise performance, and other tactics which have already been shown not to work to "fix" youthful offenders, underperforming students, and the like versus the geneticist contention (believed if not always explicitly stated) that people aren't much going to change and there's not much we can do. You are what you are, eventually.  He notes that while there is considerable persistence of personality, especially as we age, change does slowly occur.

But it did invite the question of the obvious follow up question, well, if we are changing like that over a long period of time, then is it possible to do something about it. And so we started going down that rabbit hole about 15 years ago, and we looked over the shoulder of our colleagues in clinical psychology and a paper in 2017, where the clinicians had bothered to assess personality traits along with their other dimensions that they care about, and found that seeing a therapist resulted in about a half a standard deviation of change in neuroticism, over periods of weeks, and months, as opposed to years. So, you know, one way to look at that is to say that by seeing a therapist, you get about a half of a life of change on a personality trait, which is a lot.  
He attributes the changing to environment in the long sense - the jobs we have, the places we live, the people we live with. "I was a volleyball player; if I was four inches taller, I would have been a professional volleyball player instead of a professor, because I love that sport, you know. So, you know, those types of things are not captured in the Big Five, are not captured in ability tests and the like, but are still really important parts of people's personality." Not the interventions we read about in magazines and the like.  However, for a single targeted symptom (or three!) most kinds of therapy do some good in making changes.  Not for making an introvert into an extravert, but getting someone who is very far down that scale up closer to average. He did mentione that the meta-analysis might overstate the changeability, because of publication bias.  That is, studies that showed no change are often not published, leaving only the most dramatic studies in the mix.

Speaking of intro- and extroversion, I was surprised by the mildly positive things he had to say about the MBTI, which I have panned, but he makes a good case. The problem, as with many personality tests, is with the application more than the test itself. It is not a test of maximal performance, as an IQ test would be, but of typicality, which is not only along a continuum, but probably going to follow a Gaussian distribution or Bell curve. It will therefore be split right at the point where a slight difference in score marks you as an S for Sensing versus an N for iNtuition (I know, I know). Four months later you might flip, and then flip back again.  Your own personality might merely be hovering around the midpoint, so that you aren't really either. But the oversimplification of scoring makes it look more hard-and-fast than it is. They even describe it as "four dichotomies," but anything along a continuum, especially a Bell curve isn't really that. If you just read the results correctly, paying much more attention to the more extreme scores than the middling ones, it can tell you something about yourself. The Personality Assessment Inventory, a newer instrument which I like very much, is also along a continuum in its scoring axes, but that one is explained to the subject by someone familiar with the meanings. It matters.

People like personality tests better when the negatives are removed, so if you score low on conscientiousness, you get called spontaneous rather than irresponsible. Even if you are actually irresponsible. They would use the test as a launch point to get to that understanding, rather than smacking you with it. Which is part of why I am a terrible counselor, usable only in acute, high-energy situations where "Look son, you either get this or you don't."

I liked his take on the Big Five Personality Traits as well.  He and Razib spend a good while discussing those, with examples of how a "single trait" might really be a related collection of stable traits that are not identical.

So there's different facets. And so that's why so I like to describe the Big Five as a family, loosely speaking. And so you've got things like nurturance, and agreeableness, trust, kindness, politeness, compliance, which are all different facets, within that family, they're all related. They're all positively correlated with one another, but they reflect different flavors of of the general domain that manifests differently. So please don't take my description as just the trust is the only thing there are other aspects to. And the same goes for conscientiousness. So you know, you sound like a classic academic, the high extraversion, high openness, middling conscientiousness, often because in that, that space, you've got the hard working component, that's one of the facets, but you also have the kind of conventional rule adhering component. And so typically, scientists and academics have a split on that really hard working really not conventional. And so we ended up being on the on the average, so to speak. 
 Very sensible.  Highly recommended.
 
 *You may not have access to the transcript, which is why I quoted heavily.

2 comments:

  1. MBTI makes a lot of sense to me...but I've seen claims that no one has ben able to correlate it with other personality measurements. Maybe due to the artificial binary categories rather than using a continuum approach?

    ReplyDelete
  2. Years ago, I was at a company management training class where, in addition to the sessions of finance, strategy, operations, etc we had some interesting outside speakers. One was a psych professor, a Jungian IIRC. He gave us the MBTI test and without telling us the results divided us into groups and gave us all the same task to see how we approached it. The Intuitives spent a lot of time arguing about how to approach the task, the Feeler came back in singing 'Feelings', can't remember about the others.

    Advice he gave us was: you'll be tempted to hire people who are like you in personality. Resist that temptation, or you'll all have the same blind spots and walk happily off the cliff together.

    ReplyDelete